青少年时期是习惯养成的重要时期，但现阶段我国青少年的体质健康状况却不容乐观，大量的实验结果表明体育锻炼能够能够改变人体的健康状况，但由于体育锻炼行为受多种因素共同影响，单从一方面进行干预不能让学生养成坚持锻炼的良好习惯，因此构建符合中学生的体育锻炼行为量表，从多个角度分析中学生的体育锻炼行为特点，为今后的体育改革和干预提供一定的借鉴。 本研究采用文献资料法、访谈法、逻辑分析法并基于社会生态学理论的基础上确立了《北京市中学生体育锻炼行为》初始问卷，并采用SPSS对数据进行同质性检验、探索性因素分析以及验证性因素分析得出《北京市中学生体育锻炼行为》最终量表，并通过该量表调查北京市中学生的体育锻炼行为特点。结果发现： （1）通过大量文献查阅并结合社会生态学模型理论，将影响中学生体育锻炼行为的因素划分为个人、家庭、学校、社区以及政策5个维度，在此条件下查阅文献确立了各维度下的指标，其中个人层面2个、家庭层面3个、学校层面2个、社区层面2个、政策层面2个共11个指标，并根据指标编制基于社会生态模型理论下《北京市中学生体育锻炼行为》问卷76个题项。 （2）《北京市中学生体育锻炼行为》量表初始量表包含76道题，经检验剔除57个题项，最终所剩余的19道题具有良好的区分度，总量表与各分量表的Person相关系数在0.610-0.794之间且全部项目之间为正相关，该量表的 Cronbachα系数为0.904，各分量表的信度系数在0.634-0.930之间，表明量表的信度较理想，根据验证性因子分析，《北京市中学生体育锻炼行为》量表维度划分合理且最终量表包含的19道题具有良好的区分度。 （3）经检验《北京市中学生体育锻炼行为》量表具有较高的信效度，可以作为北京市中学生体育锻炼行为影响因素调查的测量工具。 （4）不同性别中学生的体育锻炼行为在家庭和学校维度由于接受的教育资源和教育环境相似，因此差异不具有显著性，而在个人、社区、政策维度存在显著差异，同时在各个层面男生的受影响程度均大于女生，其中受政策层面的影响最大。 （5）不同学校可能由于学校的性质不同，对政策的普及和重视程度也存在一定的差异，造成不同学校间的学生在体育锻炼时政策之间的差异存在显著性；由于学校的一些规定或者对课程的调整等原因可能造成学生锻炼途径的不同选择，因此在锻炼途径的选择方面只受政策层面的影响差异存在显著性。 （6）对于不同锻炼次数的中学生在五个层面的差异都不存在显著性，同时还发现每周锻炼两次的人群受五个层面的影响都是最大的，每周锻炼一次的人群受社区和政策层面的影响最大。

Adolescence is an important period of habit, but at this stage, the physical health of teenagers is not optimistic in our country, a large number of experimental results show that physical exercise can able to change the body's health, but as a result of physical exercise behavior is influenced by many factors, from one hand to intervene cannot let students form the good habit of exercise, so the building in conformity with the physical exercise behavior scale of middle school students, from the perspectives of multiple physical exercise behavior characteristics of middle school students, for the future of the sports reform and intervention to provide certain reference. This research USES the literature material method, interview method, logic analysis, established the "Beijing physical exercise behavior of middle school students" initial questionnaire, and used SPSS to homogeneity test data, exploratory factor analysis and confirmatory factor analysis of physical exercise behavior of middle school students "" Beijing final scale, and through the questionnaire investigation of Beijing middle school students' physical exercise behavior characteristics. Results: (1) through a large number of literature review and combining the social ecology model theory, the factors that affect middle school students' physical exercise behavior is divided into individual, family, school, community, and policy in five dimensions, under the condition of the established under the various dimensions of indicators through the literature, in which the individual level 2, family level 3, the school level 2, 2 at the community level, policy level 2, a total of 11 indicators, and according to the index compiled based on the theory of the social ecological model under the "Beijing middle school students' physical exercise behavior questionnaire" 76 item. （2） The initial scale of the "Sports Exercise Behavior of Middle School Students in Beijing" contains 76 questions, and 57 items are excluded from the test. The remaining 19 questions have a good degree of discrimination, the total amount and each The Person correlation coefficient of the subscale is between 0.610-0.794 and there is a positive correlation between all items. The Cronbachα coefficient of the scale is 0.904, and the reliability coefficient of each subscale is between 0.634 and 0.930, indicating the reliability of the scale. Ideally, according to the confirmatory factor analysis, the dimension of the "Beijing middle school students' physical exercise behavior" scale is reasonable and the 19 questions contained in the final scale have a good degree of discrimination. (3) the "Beijing middle school students' physical exercise behavior" scale has a high reliability and validity, which can be used as a measurement tool for the investigation of influencing factors of Beijing middle school students' physical exercise behavior. (4) the different gender students physical exercise behavior in family and school dimensions due to accept education resources and education environment are similar, therefore has no significant difference, but significant differences in individuals, communities, policy dimensions, while the boys in all aspects of influence degree were greater than girls, which is affected by the policy level is the largest. (5) due to the different nature of schools, there may be some differences in the popularity and degree of attention paid to policies in different schools, resulting in significant differences in the policies of physical exercise among students in different schools; Because some school regulations or curriculum adjustments may cause students to choose different ways of exercise, there is a significant difference in the choice of exercise paths only under the influence of policies. (6) for middle school students with different exercise times, there was no significant difference in the five levels. Meanwhile, it was also found that those who exercised twice a week were most affected by the five levels, and those who exercised once a week were most affected by the community and policy levels.